ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue1, Iyun 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
967
967
UTILIZING ARTIFICIAL INTELLIGENCE FOR DIAGNOSING AND ASSESSING
LANGUAGE SKILLS IN EFL STUDENTS
Qodirqulova Maftunaxon Muhiddin kizi
Chirchiq Davlat Pedagogika Universiteti ingliz tili
Abstract:
Artificial Intelligence (AI) technologies are increasingly influencing educational
practices, especially in the field of language learning. This study investigates the use of AI in
diagnosing and assessing the language skills of students learning English as a Foreign
Language (EFL). The research explores how AI tools are utilized to evaluate learners’
proficiency in reading, writing, listening, and speaking. Findings indicate that AI can offer
real-time feedback, personalize learning paths, and reduce the subjectivity of traditional
assessment methods. However, limitations such as cultural bias, lack of contextual
understanding, and technical constraints remain. This paper aims to provide educators with
insights into how AI can be effectively integrated into language assessment practices.
Keywords
: Artificial Intelligence, Language Assessment, EFL, Diagnostic Tools, Language
Proficiency, Educational Technology
Introduction
As English continues to serve as the global lingua franca, the demand for accurate and
efficient language assessment tools has intensified, particularly for EFL (English as a Foreign
Language) learners. Traditional assessments often rely heavily on manual grading, which can
be time-consuming and inconsistent. In response, Artificial Intelligence (AI) has emerged as a
powerful tool that offers new possibilities for diagnosing and assessing language proficiency.
AI-based systems, such as natural language processing (NLP) and speech recognition software,
can evaluate language performance across different modalities with remarkable speed and
objectivity. However, questions remain about the reliability, validity, and fairness of such
assessments. This paper aims to explore both the potential and limitations of AI in language
skill diagnosis and assessment for EFL students.
Methodology
Research Design
This study employs a qualitative research design supplemented with a small-scale
empirical analysis. It includes a comprehensive literature review and interviews with EFL
educators who have used AI-powered tools in their classrooms. In addition, several popular
AI-based assessment platforms were evaluated based on their diagnostic capabilities.
Participants
Ten EFL teachers and thirty EFL students from three different language institutions
participated. The students were from intermediate and upper-intermediate levels, and the
teachers had experience using at least one AI assessment tool, such as Duolingo English Test,
ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue1, Iyun 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
968
968
Write & Improve (by Cambridge), or AI-based classroom apps like Grammarly and Elsa
Speak
Data Collection Tools
Interviews with teachers regarding their experiences with AI tools.
Observation of students interacting with AI-powered platforms.
Review of feedback and scoring reports generated by AI tools.
Data Analysis
The collected qualitative data were coded and analyzed thematically. AI-generated asse
ssments were compared with traditional teacher evaluations to examine consis tency and
diagnostic depth.
Result
Diagnostic Capabilities of AI Tools
The AI tools examined were effective in diagnosing grammar, syntax, pronunciation, and
vocabulary use. Platforms like Grammarly provided detailed feedback on grammatical
structure and word choice, while speech-based tools like Elsa Speak offered real-time
pronunciation analysis. These platforms were particularly useful for self-correction and learner
autonomy.
Assessment Across Language Skills
Reading
: AI was moderately effective in evaluating reading comprehension through
multiple-choice or fill-in-the-blank tasks. However, it struggled with assessing deeper
interpretive skills.
Writing
: AI systems performed strongly in identifying mechanical errors but were
weaker in assessing content coherence and originality.
Listening
: AI tools successfully tested auditory comprehension but had difficulty
adapting to varied accents and real-world audio inputs.
Speaking
: Pronunciation and fluency were effectively evaluated through speech
recognition, but intonation and spontaneous expression were often misjudged.
3.3 Teacher and Student Feedback
Teachers found AI to be a helpful assistant in formative assessment, saving time and offering
consistency. However, they stressed that AI should not replace human judgment. Students
appreciated the instant feedback but expressed concerns about over-reliance and the lack of
personalized explanations.
Discussion
Strengths of AI in Language Assessment
ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue1, Iyun 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
969
969
One of the most notable advantages of AI in EFL assessment is its ability to process and
evaluate large volumes of language data quickly and objectively. This is especially beneficial
in large classrooms where individualized feedback is challenging. AI also supports
personalized learning by identifying specific areas of weakness and tailoring practice tasks
accordingly.
Limitations and Challenges
Despite its benefits, AI assessment tools face notable limitations:
Contextual Understanding
: AI often fails to grasp context, tone, or culturally
nuanced language use.
Bias
: Training data may reflect native speaker norms, which can disadvantage EFL
learners from diverse backgrounds.
Technical Limitations
: Inconsistent internet connectivity, device incompatibility, and
software glitches can hinder assessment accuracy.
Implications for EFL Education
The integration of AI into language assessment should be viewed as a complementary
approach rather than a complete replacement for human assessment. Educators should be
trained to interpret AI feedback and guide learners in making sense of automated results.
Policy-makers must ensure that AI tools are used ethically and inclusively, with regular
validation against pedagogical standards.
Conclusion
Artificial Intelligence holds significant promise for diagnosing and assessing language
skills in EFL contexts. Its speed, scalability, and analytical capabilities make it an ideal
supplement to traditional assessments. However, AI tools are not without flaws. Limitations in
understanding context, evaluating creativity, and ensuring equity mean that human oversight
remains essential. When implemented thoughtfully, AI can enhance language education by
providing both learners and educators with deeper insights into the learning process.
References
1.
Burstein, J., & Chodorow, M. (2010). Automated essay scoring for nonnative English
speakers.
AI Magazine
, 31(3), 27-36.
2.
Chapelle, C. A., & Voss, E. (2016). Evaluation of language learning technologies.
International Journal of Artificial Intelligence in Education
, 26(1), 27-36.
3.
Li, Z., Link, S., & Hegelheimer, V. (2021). The role of AI in EFL writing assessment.
Language Testing
, 38(4), 491–512.
4.
Ranalli, J. (2020). Automated written corrective feedback: What do learners make of it?
Computer Assisted Language Learning
, 33(3), 205–231.
5.
Wang, Y., & Yu, S. (2022). AI-assisted speaking assessment: Potential and pitfalls.
Assessing Writing
, 52, 100591.
6.
Xu, Y., & Brown, G. T. L. (2016). Teacher assessment literacy in practice: A
reconceptualization.
Teaching and Teacher Education
, 58, 149–162.
ISSN: 3030-3931, Impact factor: 7,241
Volume 8, issue1, Iyun 2025
https://worldlyjournals.com/index.php/Yangiizlanuvchi
worldly knowledge
OAK Index bazalari :
research gate, research bib.
Qo’shimcha index bazalari:
zenodo, open aire. google scholar.
Original article
970
970
7.
Duolingo. (2023). Duolingo English Test: Research and development. Retrieved from
[https://englishtest.duolingo.com/research]
8.
Cambridge University Press. (2022). Write & Improve with Cambridge English.
Retrieved from [
